APARAJIT Micro AI Accelerator IP

APARAJIT FAMILY ENTRY-LEVEL CONFIGURATION 1 + 1 CORE

Compact AI Accelerator IP for Embedded AI and Intelligent Computing

Built for Compact AI Acceleration

APARAJIT Micro is the entry-level configuration in the APARAJIT AI Accelerator IP family. It combines one AJIT Processor Core with one APARA AI Accelerator Core, providing a compact semiconductor platform for AI inference, tensor processing, and embedded computing applications.

Built on the common APARAJIT architecture, APARAJIT Micro shares the same hardware and software architecture as the rest of the APARAJIT product family. This enables engineering teams to develop applications on a common platform while scaling to larger APARAJIT configurations as application requirements evolve.

APARAJIT Micro is designed for organisations developing embedded AI platforms, robotics, IoT inferencing, and signal processing applications requiring programmable processing together with dedicated AI acceleration.

1 + 1

AJIT CORE + APARA AI CORE

5-Const.

GNSS SUPPORT

256 Ops/Cyc

PEAK (FP4 / INT4)

4 × MIMD

INDEPENDENT THREADS

256-bit

LOAD / STORE WIDTH

– Why APARAJIT Micro?

Compact acceleration, common architecture.

Compact AI-enabled semiconductor platforms require programmable processing together with dedicated AI acceleration within an efficient System-on-Chip (SoC) architecture.

APARAJIT Micro combines one AJIT Processor Core with one APARA AI Accelerator Core to provide programmable processing and AI acceleration within a common semiconductor architecture.

As part of the APARAJIT product family, APARAJIT Micro maintains software portability across Micro, Mini, and Mega configurations, enabling engineering teams to scale product development while maintaining a consistent development workflow.

APARAJIT Micro is designed for organisations developing embedded AI platforms, robotics, IoT inferencing, and signal processing applications requiring programmable processing together with dedicated AI acceleration.

Specification APARAJIT Micro
AJIT Processor Cores 1
APARA AI Accelerator Cores 1
GNSS Support 5-Constellation GNSS
– Technical specifications

APARA AI Accelerator performance and architecture.

Compute Performance

FP4 / INT4256 Ops/Cycle
FP8 / INT8128 Ops/Cycle
FP16 / INT1664 Ops/Cycle
FP32 / INT3216 Ops/Cycle
FP64 / INT648 Ops/Cycle

APARA Architecture

Execution Model4 Independent MIMD Threads
Hardware Execution Lanes8
Instruction Cache16 KB Bundled Instruction Cache
Memory Banks4 Parallel Memory Banks
DMADedicated DMA Engine
Load / Store Width256-bit
– Architecture overview

One unified semiconductor architecture.

APARAJIT Micro combines one AJIT Processor Core with one APARA AI Accelerator Core within a unified semiconductor architecture for embedded AI and intelligent computing.

The APARA AI Accelerator architecture incorporates four independent MIMD threads, eight hardware execution lanes, four parallel memory banks, a dedicated DMA engine, a bundled 16 KB instruction cache, and a 256-bit load/store interface. It supports AI inference, tensor processing, and parallel computing across FP4/INT4, FP8/INT8, FP16/INT16, FP32/INT32, and FP64/INT64 precision formats.

As a member of the APARAJIT AI Accelerator IP family, APARAJIT Micro shares a common hardware and software architecture with APARAJIT Mini and APARAJIT Mega, enabling software portability and scalable product development across deployment configurations.

– Key features

What APARAJIT Micro provides.

One AJIT Processor Core

Integrates one AJIT Processor Core to provide programmable processing and system control within the APARAJIT architecture.

One APARA AI Accelerator Core

Includes one APARA AI Accelerator Core designed for AI inference, tensor processing, and parallel computing workloads

Common APARAJIT Architecture

Shares the same hardware and software architecture as APARAJIT Mini and APARAJIT Mega, enabling software portability across the product family.

Integrated Memory Architecture

Four parallel memory banks, a dedicated DMA engine, a bundled 16 KB instruction cache, and a 256-bit load/store interface for efficient data movement.

Multiple Precision Support

Supports FP4/INT4, FP8/INT8, FP16/INT16, FP32/INT32, and FP64/INT64 precision formats for AI inference, tensor processing, and parallel computing.

– Applications

Where APARAJIT Micro fits.

Drones

Provides AI acceleration for embedded drone platforms requiring onboard AI inference and intelligent processing.

Robotics

Supports robotics applications requiring programmable processing together with dedicated AI acceleration.

IoT Inferencing

Enables AI inference on embedded IoT devices performing local intelligent processing.

Signal Processing

Accelerates signal processing workloads across embedded and intelligent semiconductor platforms.

– Why choose APARAJIT Micro

Why Choose APARAJIT Micro?

  • One AJIT Processor Core
  • 5-Constellation GNSS Support
  • Software Portability Across the APARAJIT Product Family
  • Licensable Semiconductor IP for Custom System-on-Chip (SoC) Integration
  • One APARA AI Accelerator Core
  • Common APARAJIT Architecture
  • Multiple Floating-Point and Integer Precision Formats

– Inside the stack

Explore the Complete Architecture

APARAJIT AI Accelerator IP provides dedicated AI acceleration for intelligent semiconductor platforms, while AJIT Processor Core IP delivers programmable processing, system control, and System-on-Chip (SoC) management.

Together, they form the Powai Labs semiconductor IP portfolio, enabling intelligent computing across embedded systems, robotics, industrial automation, defence, aerospace, navigation technologies, and AI-enabled semiconductor platforms.

Request Product Information
Enquire Now